Combining Bayesian and Evidential Uncertainty Quantification for Improved Bioactivity Modeling.

ABSTRACT: Uncertainty quantification (UQ) has been recognized as a prerequisite for reliable and trustworthy computational modeling in drug discovery. Two widely considered paradigms, Bayesian methods (deep ensemble and MC dropout) and evidential learning, differ in their computational demands and expressivity of uncertainties, excelling in complementary settings. Here, we propose hybrid approaches that combine both paradigms and benchmark them on the Papyrus++ data set across two end points (xC₅₀, K_x) and multiple split strategies. Our ensemble of evidential models (EOE) consistently achieves the best overall performance, yielding the lowest RMSE and leading CRPS and interval scores, including under the most challenging distributional shifts. While large e

SUBMITTER: Khalil B 

PROVIDER: S-EPMC12728939 | biostudies-literature | 2025 Dec

REPOSITORIES: biostudies-literature